One of the central problems in modern human genetics is to understand the functional impact of genetic variation. Of the millions of DNA positions that vary among humans, which sites actually impact human phenotypes and disease? It is becoming increasingly clear that noncoding variants that impact gene regulation play central roles in the genetics of disease, yet these variants remain difficult to interpret. Inthis project we will use rheumatoid arthritis as a model system for tackling these problems, using a combination of computational and experimental techniques. On the computational side, we will develop statistical approaches for integrating functional information from genome-wide assays with data from GWAS studies for complex traits. Our work will allow us to infer which cell-types are most important for any given disease, to improve mapping power, and to improve our ability to identify the most likely causal variants. We also propose novel approaches to validation, based on measuring cellular phenotypes in sorted immune cells and performing genome editing of specific sites. We anticipate that our work will lead to detection and validation of novel RA loci, as well as new techniques of general utility in this field.